[vigouroustester]

With the stochastic nature of LLM’s and the ever moving goal-posts, a framework not based off of knowledge that might already be in its memory is definitely needed